Diagnostic study, design and implementation of an integrated model of care in France: a bottom-up process with continuous leadership.

BACKGROUND: Sustaining integrated care is difficult, in large part because of problems encountered securing the participation of health care and social service professionals and, in particular, general practitioners (GPs).
PURPOSE: To present an innovative bottom-up and pragmatic strategy used to implement a new integrated care model in France for community-dwelling elderly people with complex needs.
RESULTS: In the first step, a diagnostic study was conducted with face-to-face interviews to gather data on current practices from a sample of health and social stakeholders working with elderly people. In the second step, an integrated care model called Coordination Personnes Agées (COPA) was designed by the same major stakeholders in order to define its detailed characteristics based on the local context. In the third step, the model was implemented in two phases: adoption and maintenance. This strategy was carried out by a continuous and flexible leadership throughout the process, initially with a mixed leadership (clinician and researcher) followed by a double one (clinician and managers of services) in the implementation phase.
CONCLUSION: The implementation of this bottom-up and pragmatic strategy relied on establishing a collaborative dynamic among health and social stakeholders. This enhanced their involvement throughout the implementation phase, particularly among the GPs, and allowed them to support the change practices and services arrangements.

Introduction

In health systems, improving and reorganizing elderly care has become a priority in order to cope with the challenges inherent in meeting the needs of older persons [1, 2]. Projects implementing integrated care have taken centre stage as a way to improve quality and efficiency in care for the elderly [3, 5]. But adoption of these integrated care models generates implementation problems, and it is difficult to secure the participation of professionals, particularly that of general practitioners (GPs) [6-8]. While GP participation represents one of the key success factors in the implementation of integrated care models [9, 11], research has found a lag between the conceptual approach of these models and professional practice in primary care [12, 13]. Implementations need to be founded in a strong leadership that encourages the participation of professionals and prevents resistance behaviours [14]. But the role of these leaders has not been clearly defined [15] and most of studies were in the US and also in hospitals [16].

Since the health system in France is fragmented in many ways, it appeared to be a good setting for the implementation of a model of integrated care. Strong fragmentations exist between medical services and social services, community-based and hospital-based services, healthcare professionals and family caregivers, as well as long-term and acute care [17]. Over the last 30 years, every top-down attempt at reorganizing services made by central authorities, has coordinated gerontology services and established networking in which professionals participate voluntarily [18]. These attempts featured no formalized collaboration, a lack of participation by professionals, and strong competition between care providers (no limitations on their areas of competency) which have reinforced fragmentations [19].

A new strategy is, therefore, needed for designing and implementing an integrated care model that will foster adoption by professionals and changes in practices. Such methods involve carrying out a diagnostic study to understand current practices and having professionals engaged in the design process [14, 20] so that the intervention meets professionals' expectations and takes their local working context into consideration [12, 21, 22]. To address this problem, an integrated care model for community-dwelling elderly people was developed and implemented in France, using an innovative bottom-up strategy based on legitimate leadership. The aim of this article is to present this strategy, which includes a series of distinct steps and a continuous and flexible leadership. This approach was taken in order to foster the professional participation and to support change practices and services arrangements for elderly people with complex needs.

A bottom-up and pragmatic process

This process had three distinct phases (see Figure 1).

Figure 1.

Steps in the integrated care process.

First step: a diagnostic phase

The objective of the diagnostic study was to describe the comprehensiveness of the continuity of care and the current coverage of professional practices and service arrangements.

Sampling

Participants were selected first through a purposeful sampling strategy [23]. All professionals responsible for delivering community-based and hospital-based services for older people in an area of Paris, were identified. In addition, independent fee-for-service professionals, including GPs with old patients, were identified from lists of persons receiving these services. The sample was completed with a snowball sampling strategy: names of other potential participants were collected from the initial group in order to complete the sample and ensure good representation of professionals. All the stakeholders were contacted and asked to participate in this building phase of the integrated care model. Only two participants, GPs, refused to be involved. The final sample included representatives from all the hospitals (one public general hospital, one public geriatric hospital and one non-profit geriatric hospital) and community-based services (non-profit nursing services, private home services, public social services) as well as independent professionals. In all, the sample comprised 56 stakeholders: 29 professionals working in primary care including 10 independent professionals (8 GPs, 1 nurse, 1 physiotherapist) and 19 professionals from community-based services (4 nurses, 2 auxiliary nurses, 6 social workers, 2 homecare workers, 5 managers), 23 professionals working in hospitals (3 geriatricians, 2 psychiatrists, 2 emergency physicians, 4 nurses, 3 physiotherapists, 5 social workers, 4 managers) and four managers from funding agencies.

Data collection

The data were collected in 45-minute individual face-to-face interviews at a professional workplace using a semi-structured interview guide. These interviews explored current practices, perceived issues and expectations regarding the care of elderly people. The interviews were complemented by observation and documentation to enhance the validity of the data. The observations were made at various organizations (hospitals, community-based health and social service centres), and documents (minutes, memos, activity reports) were also collected from each setting and analyzed.

Results

The diagnostic study was undertaken to identify professional practices based on ‘who does what,’ and it revealed overlaps and gaps. It was used to reveal positive points as well as dysfunctional aspects of care paths that were leading to adverse outcomes for elderly persons. The results of this diagnostic study have been presented in another article [24].

Second step: design of the COPA model (Coordination Personnes Agées)

The same stakeholders were used to build the theoretical model based on the improvement areas in function of the diagnostic phase.

Focus group

Four focus groups met in 90-minute sessions, held in parallel, and each group met four times. With the exception of the GPs, all the participants were distributed among three of the four groups, balancing professionals from the social and health services and from primary care and hospital settings. The fourth group consisted of the GPs only, who participated at lunch meetings in order to foster their participation without providing financial compensation.

A three-step process

First, findings from the diagnostic study were presented to focus groups and stakeholders were asked to express their expectations concerning the reorganization of services and change practices.

Second, investigators presented a framework of six critical dimensions of integrated care based on the literature: (1) a single entry point, (2) a target population (3) a primary care management process, (4) a transition between primary and secondary care, (5) a communication system and (6) a structured governance. The framework was used as a guide to facilitate discussions. Stakeholders in the focus groups were asked to reach a consensus on the detailed characteristics of a new integrated model based on the dimensions of this framework.

Third, conclusions from the focus groups with objectives and characteristics of the integrated care model were presented to all stakeholders for discussion and validation, leading to a consensus.

Characteristics of the COPA model

There were two objectives for the integrated care model. First, the model needed to systematize coordination at the clinical and service levels, with a better fit between social and health professionals in order to respond to the population's needs. Second, COPA would change services organisation by decreasing emergency room visits and hospital and nursing home admissions.

To attain these objectives, the resulting model has six features: (1) a single entry point defined as a process with standardized and converging processes for directing elderly people. Persons with complex needs are directed toward case management and ineligible ones are referred to community-based services (a simple referral with no follow-up); (2) the target population for case management is selected on the basis of the complexity of their health and social conditions. The criteria were developed based on an existing tool (the InterRai Contact Assessment) with seven items: four items on difficulties with ADL (activity of daily living), one item on cognitive deficiency, one item on poor perceived health and one on shortness of breath [25]. Two social items suggested by the focus groups, living alone and the unavailability of a caregiver, were also added; (3) the case management process was developed taking into account existing clinical coordination practices for the elderly in order to avoid duplications. The case management is considered intensive, with full-time nurses acting as case managers, each of whom is responsible for monitoring a case load of 40 persons with complex needs. The role of the GP is defined as a professional who works closely with the case manager in a two-person team. Each case manager works with a limited number of GPs and each of the GP's patients with complex needs are assigned to the same case manager, producing a high patient/GP ratio per case manager; (4) the transition between primary and secondary care is seen within the larger context of arrangements between independent professionals and community-based health and social services, whether private and public. A multidisciplinary ‘support team’ supports the new services arrangements and the changes in professional practices. The support team consists of two geriatricians from public and private hospitals, a psychologist from a community-based health services and a coordinator from a nearby social service agency. The members of this support team spend half their time working in their departments and the other half of their time participating in case management; (5) the communication system; and (6) a non-profit consortium will be created to manage public funding and administrative tasks. The consortium consists of all the managers of services participating in the building of this integrated care model.

A detailed description of the COPA model of integrated care has been provided in an earlier paper [26].

Third step: implementation

Following the diagnostic study and the model design process, the integrated care model was implemented. It was a two-phase implementation, involving an adoption phase, in which the theoretical model was transformed into a practical model, and a consolidation phase, in which operations came to resemble routines and the participation of professionals stabilized. In the course of the implementation, new practices and services arrangements reached professionals that had not participated in the diagnostic and design phases.

Stage 1: the adoption phase

Case managers (3 professionals) and members of the support team were recruited and deployed in primary care. These professionals were trained in the case management process and in performing a comprehensive assessment using the InterRai Minimum Data Set for Home Care (MDS-HC) [27, 28]. The modes of intervention used by the case managers and the support team were refined. Then plenary sessions were held to inform professionals and services in the territory about the integrated care model. The exception was GPs, who were informed in personal visits (modelled on the visits made by pharmaceutical representatives). These information meetings provided an opportunity for each professional to meet case managers and the support team in person. Case managers started working with the GPs who had participated in the building of the model. Inclusion of elderly persons then began, so that case loads gradually grew by a maximum of 10 new persons per month per case manager.

This phase provided an opportunity to make adjustments of the single entry point, from the request to the interventions chosen. Finally, the non-profit consortium was created to manage the public funding.

Stage 2: the maintenance phase

The maintenance phase coincides with the systemization of the coordination around persons with complex needs and coordination between and within services. This consolidation phase was used to disseminate the model across the given area.

The single entry point informs and refers elderly people to services on the basis of their needs. By connecting the services, all requests can be processed in the same manner, and the function charged with directing elderly persons, is able to simplify the care paths.

As for persons living at home, the case manager informs, advises, supports and monitors their welfare. In terms of services, the case manager coordinates the actions of all professionals (social and health professionals) and informal caregivers. The case manager may act as mediator or even advocate between the person and the services. The number of GPs participating in case management grew to an average of 10 GPs per case manager. The support team intervenes at the request of case managers and other professionals working in community-based services who are having difficulties with an elderly person living at home. This means that the geriatrician provides specific geriatric expertise (such as on behavioural disorders, memory loss, bedsores, palliative care, etc.). The geriatrician organizes the direct hospitalizations without emergency room visits. The psychologist intervenes with elderly people, but also with the professionals and caregivers. The psychologist supports also case managers. The coordinator from a nearby social service agency monitors people with less complex needs.

A shared communications system is still in development because of incompatibilities between existing tools used by the participating services.

Governance of the model is provided by the non-profit consortium with the participation of managers from community-based services and hospital settings.

Continuous evaluation

The bottom-up approach was continually evaluated throughout the entire length of the process. The evaluation provided information on the participation of stakeholders in the diagnostic and design phases, demonstrating the roles played by local context and leadership. In the course of the implementation, the on-going evaluation was both qualitative and quantitative. The qualitative study comprised case studies of professional practices and services arrangements produced by the implementation of the integrated care model. The goal of the quantitative study was to analyze the impact of interventions in terms of quality of care and services utilization.

The integration process took place under a continuous and flexible leadership to support change in all the stages, from the diagnostic to implementation of the model of care (see Table 1).

Table 1.

Leadership and management

	Diagnostic	Design	Adoption		Maintenance
Length (years)	1	1	2		1
Level of leadership and location	Clinician (hospital) and researcher (research group)	Clinician (hospital) and researcher (research group)	Clinician (hospital)	Managers (services)	Support team (clinical and services)	Managers (services)
Training	Accustomed to integrated care and geriatric practices	Accustomed to integrated care and geriatric practices	Geriatrician	Managers participating in the process	Multidisciplinary team	Managers participating in the process
Implementation of formal processes	Individual leading with face to face interviews	Group leading with focus group	Case management meeting one a week	Managers meeting one a month	Case management meeting one a week	Managers meeting two a year
Leaders' role	• Selection and recruitment of stakeholders• Record professional practices• Analyse data• Create trust	• Encourage to develop solutions• Create a sense of being part of a group• Assume the leadership role• Build links with research	• Vigilance in protecting the process• Systematize coordination• Legitimate the case manager• Broaden the professionals commitment	• Vigilance in protecting the process• Develop arrangements between services• Limit competition between public and private services• Governance the non-profit consortium	• Support the case management• Reinforce inter-professional bonds• Disseminate good practices• Change in service arrangements	• Develop arrangements between services• Governance the non-profit consortium• Prevent the model from becoming too rigid
Incentives and sources of legitimacy in leadership	• Hospital setting and clinician practices• Selected sample of stakeholders• Representative of the stakeholders• Researcher background	• Realisation the diagnostic study• Structuring with a framework• Process with focus group• Having time for building	• Knowledge of professionals• Clear definition of the complexity• Focus on population with complex need• Weekly case management meeting	• Managers having participated in the process• Interdependence between services• Support team	• Double membership• Multidisciplinary team	• Double leadership
Strategy with regard to GPs	Selected GPs	Keeping them informed	• GP is responsible for decision-making• Case manager responsible for time-consuming tasks	GPs participate	• Advice from geriatrician• Direct hospitalization

Leadership with clinician-researcher for the diagnostic and design phases

The diagnostic and design phases took two years and were monitored by two investigators who are physicians. The leadership included a geriatrician (MDS) from the public hospital who has a clinical background and a public health physician with a researcher background (IV). These physicians are accustomed to international models of integrated care.

The diagnostic phase

In the diagnostic phase, the investigators identified and recruited all stakeholders in the integrated care process. They provided a year of one-on-one support in order to record perceptions of professional practices and to understand relationships and underlying issues. The data were analyzed using the grounded theory-building approach described by Pandit [29]. This theory was used to identify relevant categories and the creation of relationships and to develop the conceptual framework that emerged from the data.

Among the incentives for the leadership, the fact that the clinician was associated with the hospital reinforced the participation of community-based professionals interested in having ‘open’ hospitals. The leadership's experience in clinical practice and its knowledge of home interventions for elderly, contributed to establish trust with the stakeholders. Selecting participants on the basis of their activities with the elderly made it possible to bring on board those professionals with the greatest motivation to imagine changes in services organization. The initial selection of GPs whose patients were largely elderly people (>70%) was effective in terms of obtaining their commitment. No time was lost trying to recruit a large sample of GPs in the territory. The sample of participants was also representative of all professionals who provide services in the home, particularly social workers and home helpers to have a real diagnostic of practices in this area. The research background of the leadership allowed conducting this phase.

Design phase

During the design phase, the investigators encouraged stakeholders to develop solutions based on the results of the diagnostic study, particularly in terms of the strengths and weaknesses of current elderly care. The investigators provided support at each focus group, and after each meeting they prepared a summary of the progress made in developing integrated care. When contradictions became apparent, the investigators presented them at the following focus group so that the partners could clarify their points of view. The social and health professionals learned from each other, and this reinforced their interdependencies, developed good lines of communication and created common values. Throughout this phase, the leaders ensured that no professional or group of professionals became too powerful within the group dynamic and leaders assumed the leadership role. The leaders checked also that all managers had a clear mandate from their services to take decision-making during this design phase. The focus groups reinforced interactions between the leaders and the stakeholders and allowed building links between research and stakeholders.

In terms of incentives, the diagnostic study made stakeholders more aware of the problems being addressed. The conceptual framework helped them to initiate the required changes and to structure the development of the new model. Focus group involved together clinicians and managers and contributed to understand the other's professionalism, which is key for gaining their commitment. The focus group approach was also used so that stakeholders have had the time they needed to become part of the process.

Finally, the special GP focus group was organized around the time constraints of these very busy professionals. This approach kept them informed throughout the development phase and took their points of view into account, enhancing the new integrated care model.

Leadership with clinician and service managers for the implementation phase

During the implementation phase, the leadership was divided between the clinical level and the service level and the same clinician was involved throughout the process. Implementation is a fragile phase for stakeholders who are learning how to ‘work together.’ This is why the clinical and service leadership must remain sensitive to this issue and not arrive at hasty judgments, but rather try to always maintain trusting relationships.

At the clinical level

The same clinician leader monitored the adoption phase in concert with the case managers and the support team. The weekly meeting was used to analyze how the single entry point was working in terms of the relevance of directing elderly people. Case managers would discuss problems they were having coordinating their interventions with professionals in community-based or hospital services and with independent professionals. Based on these case studies, the leader clarified dysfunctional areas and revised task distributions, refining and legitimating the role of the case manager. The clinician leader ensured that the case manager not always calling on the same professional to provide care and rather, he sought to broaden the commitment of different health and social services for the case management.

In terms of incentives for the leadership, the fact that the clinician leader knew professionals in this area was a great asset persuading them to change their practices. A clear definition of the complexity using a binary tool, simplified the direction of the population. Focusing on elderly persons with complex needs helped to foster the participation of health and social professionals. Such conditions had previously thwarted the best efforts of each professional, and the presence of a case manager proved helpful to everyone involved. The weekly discussions of the case management interventions were as a way to structure the adoption phase and make adjustments to the implementation.

Finally, since a ‘physician speaks to a physician,’ the clinician leader supported GPs throughout this stage and initiated the two-person team and explaining them the attraction in working with the case manager. This clinician was able to transform the classic hierarchical relationship between a GP and a nurse into a more balanced relationship between a GP and a case manager. GPs are responsible for medical decisions and share clinical responsibility with the case managers, who assume a wide range of tasks, relieving the GPs of time-consuming tasks.

In the services

Leadership is provided by the managers who have participated in the previous phases. The position at the head is held alternately by a representative of a community-based service and a representative of a hospital-based service. This group met monthly during the adoption phase and every six months in the maintenance phase. GPs were included in this group. This managerial leadership group discusses arrangements between and within services concerning new interfaces, and manages funding for the case managers and the support team. The group has had to address problems arising from the competition between public and private community and hospital-based services. Some rivalry arose, since each wanted to take a leadership role and have case managers ‘under their control’.

In terms of incentives, the leadership relied on the interdependence of services witch provided from the ‘work together’ during the building of the integrated care. It suggests the beginning of joint responsibility as part of this re-organization, with a population-based and public health approach (meeting needs and eliminating redundancies). The leadership rotation caused some instability in the implementation phase, since changes of governance occurred too frequently. The support team provided information about the clinical level and contributed to help for decision-making and for managing the change.

During the maintenance phase, the managers leadership group met twice a year. The leadership continued to develop new services arrangements, governed the non-profit consortium and prevented the model from becoming too rigid. A double leadership was the main condition to maintain this work in progress with links between clinical and service level.

Between the clinical and service levels

The support team intervened in the leadership of the maintenance phase and gradually assumed the responsibilities of the clinician leader in two ways. First, at the clinical level, the team supported case management, initiated relationships with new professionals and provided interdisciplinary training. The geriatrician and the psychologist made recommendations and they disseminated good practices. Second, at the service level, the geriatrician was involved in opening lines of communication with specialized services; here, more time was required to convince the hospital physicians to strengthen their bonds with primary care professionals. The presence of two geriatricians from public and private hospitals served to strengthen links between these two structures.

The incentives for leadership provided in the double membership, with ‘one foot in a service and the other in case management.’ The multidisciplinary team was also founded on a win-win relationship with professionals. For example, a social worker who participates in an intervention for an elderly person at the request of a case manager, may ask the psychologist's intervention for another case.

Members of the support team were always working with the express agreement of the GP. The GP asked an advice for one of his patient from the geriatrician who made recommendations but never prescriptions. The geriatrician organized the hospitalizations.

Conclusions

Several lessons have been learned from this experiment. The bottom-up process makes it possible to account for local fragmentation and respond by building a model of integrated care. We decided not to implement a model of care and adapt it to the given area. Rather, our approach makes it possible to involve professionals who are working in the field to build their own model of care, to address their needs and to foster their participation in the process [30]. This approach did not take the macro level into account; there were no decision-makers or funding authorities involved in the process because French system does not currently have a single source of funding for health and social services. This pragmatic process was based on a selected sample of stakeholders who were the most likely to build the model. The process was given enough time and could be found by following the ‘classic’ Plan-Do-Check-Act (PDCA) improvement process, an iterative four-step problem-solving process. This process made it possible for stakeholders to understand that they alone did not hold part of the solution, rather it was only together that they could develop a solution [31]. Time is also needed to systematize coordination and case management should not be implemented without rearranging services and changing practices. Moreover, the involvement of the GPs was handled separately throughout the building of the model, and this helped to secure their participation. It allowed them to feel associated with the process, without the impression that the process was being decided on a top-down basis.

Leadership was the second key aspect of this innovative strategy. The leaders must be legitimate, continuous and flexible throughout the entire process. The strong clinical dimension of the leadership (the geriatrician and the support team) allowed changing the day-to-day decisions of stakeholders and dealing to influence decision-making. The implementation of an integrated care model for community-dwelling elderly persons was also facilitated by the fact that part of the leadership came from the hospital. It helped to avoid acute-care driven approaches, which are often seen in France in a very hospital-centred system. The existence of continuous leadership throughout the process supported stakeholders from the theoretical construction phase to the application of the model. The leadership was also tailored to the various phases of the process. The coupling clinician and researcher supported the stakeholders' model-building and the coupling clinician and managers reinforced the coordination of the service level and the governance. Finally, a transfer of responsibilities from the clinician leader to the support team was organized to facilitate the dissemination of the model and avoiding its personification.

This innovative strategy concerns both a process and a leadership. The leadership is supported by the process, and vice versa, such that it is difficult to imagine managing this kind of change without one or the other. What is important to remember for future applications is that this strategy maintains the in-progress character of the integration and its sustainability. Time is also required; changes to practices do not occur in a big bang, but rather by being steady supported through the different stages [32]. In addition, some questions remain, particularly with respect to leadership, its exact role, how it is constituted, the level of responsibility it carries and the training required by the leaders so that they will be seen as legitimate [15]. Selecting someone from the inner circle would appear to present advantages, given that their knowledge of the issues and the clinical dimension provide additional means for persuading colleagues that things should be done differently. Beyond the choice of a leader, the interactions between the stakeholders and the leader is also key for fostering the professional participation and responding to population needs [33].